One-dimensional Cellular Automaton Simulated Model for Reproducing Uncertainty of Traffic Flow

In existing one-dimensional cellular automaton models of traffic flow, drivers’ varying behaviors are described by the deceleration probability. However, many interrelated factors contribute to the wide range of driving behaviors observed, such as behavioral habits and thought patterns of individual drivers, differences in the vehicles’ capabilities and the actual traffic environments etc.. It’s a phenomenon of great uncertainty. A new one-dimensional cellular automaton model which reproduces the uncertainty of traffic flow was described. The deceleration probability in the new model does not obey a certain random distribution but follows a broad general distribution---the Erlang distribution. In the process of the simulation, the parameter values of the Erlang distribution are also changed randomly, which describes the real traffic state more accurately. Simulations based on the new model were performed. Compared with the simulated results from other such models, the results more closely approximate real traffic flow, and the new model describes the transition process from free traffic flow to traffic jams more precisely.